Learn R Programming

live: Local Interpretable (Model-agnostic) Visual Explanations

Installation

To get started, install stable CRAN version:

install.packages("live")

or the development version:

devtools::install_github("ModelOriented/live")

See the latest changes.

Features coming up next:

  • better support for comparing explanations for different models / different instances,

  • improved Shiny application (see live_shiny function in development version).

If you have any bug reports, feature requests or ideas to improve the methodology, feel free to leave an issue.

Materials

Find the paper about live and breakDown in R Journal.

Website: https://mi2datalab.github.io/live/

Conference talks on live: Wrocław 2018, Berlin 2017.

Python implementation of LIME and info about the method: https://github.com/marcotcr/lime

Copy Link

Version

Install

install.packages('live')

Monthly Downloads

251

Version

1.5.13

License

MIT + file LICENSE

Issues

1

Pull Requests

0

Stars

35

Forks

5

Maintainer

Mateusz Staniak

Last Published

January 15th, 2020

Functions in live (1.5.13)

plot.live_explainer

Plotting white box models.
wine

Red wine characteristics and quality.
sample_locally

Generate dataset for local exploration.
print.live_explorer

Generic print function for class live_explorer
print.local_permutation_importance

Print method for local_permutation_importance class
plot.local_permutation_importance

Plot local permutation importance
print.live_explainer

Generic print function for live explainer
identity_kernel

LIME kernel that treats all observations as equally similar to observation of interest.
local_approximation

Fit local model around the observation: shortcut for DALEX explainer objects
gaussian_kernel

LIME kernel from the original article with sigma = 1.
euclidean_kernel

LIME kernel equal to the inverse of euclidean distance.
fit_explanation

Fit white box model to the simulated data.
local_permutation_importance

Local permutation variable importance
live_shiny

Function that starts a Shiny app which helps use LIVE.
add_predictions

Add black box predictions to generated dataset
live

live: visualizing interpretable models to explain black box models.